w; nd
Subject: Re: [PATCH] D25343: [OpenCL] Mark group functions as convergent in
opencl-c.h
Hello,
> As far as I understand the whole problem is that the optimized functions are
> marked by __attribute__((pure)). If the attribute is removed from your
> example, we get LLVM dump preservi
Hello,
> As far as I understand the whole problem is that the optimized functions are
> marked by __attribute__((pure)). If the attribute is removed from your
> example, we get LLVM dump preserving correctness:
>
> define i32 @bar(i32 %x) local_unnamed_addr #0 {
> entry:
> %call = tail call i3
Just my two cents.
Let's consider a simple example:
for (int I = 0; I < 2; I++) {
A = Compute();
barrier();
Use(A);
}
This does not mean Compute() needs to be executed for all the loop iterations
before barrier() being executed. This only means that during each loop
iteration, Compute()
l.com; Clang
Commits; Sumner, Brian; Stellard, Thomas; Arsenault, Matthew
Subject: RE: [PATCH] D25343: [OpenCL] Mark group functions as convergent in
opencl-c.h
+ Tom Matt
Thanks Ettore.
I think OpenCL is subject to the same issue, and noduplicate does not help
either.
Basically if a funct
, Yaxun (Sam)
Cc: Ettore Speziale ; alexey.ba...@intel.com;
anastasia.stul...@arm.com; aaron.ball...@gmail.com; Clang Commits
; Sumner, Brian
Subject: Re: [PATCH] D25343: [OpenCL] Mark group functions as convergent in
opencl-c.h
Hello guys,
>> Should we deprecate noduplicate then as conv
Hello guys,
>> Should we deprecate noduplicate then as convergent should cover both use
>> cases for OpenCL I believe? As far as I understand noduplicate was added
>> specifically for SPMD use cases...
>
> noduplicate has different semantics than convergent. Although it is proposed
> for SPMD
